Study of generalized magneto-optical ellipsometry measurement reliability

We present an experimental and theoretical study of the reliability of generalized magneto-optical ellipsometry measurements, investigate its dependence from data set acquisition geometries, as well as investigate the underlying physics of light reflection for magneto-optical materials to explain the observed behavior. Specifically, we compare the use of two different grids of data points and evaluate their reliability and repeatability in a comparative study. We find that the conventionally used square grid is actually not ideal for generalized magneto-optical ellipsometry (GME) measurements and that the also investigated diagonal lattice is clearly superior. These experimental results were reproduced in theoretical calculations of the detection process. The physical origin of this behavior was identified to be related to the "quality" of the individual data points that are included in the data analysis process, with the highest quality data being achieved near the crossed polarizer line, i.e., the region that is more prominently utilized in the diagonal grid approach presented here. These results will help to improve the precision and the data acquisition time of GME measurements. (C) 2012 American Institute of Physics. [http://dx.doi.org/10.1063/1.4720471)